Hyperion Partners L.P. and Bank United Corp. (collectively “Hyperion”), and Bank United appeal the decision of the United States Court of Federal Claims granting them only part of the damages they sought for the government’s breach of several agreements caused by the enactment of the Financial Institutions Reform, Recovery, and Enforcement Act of 1989 (“FIRREA”) Pub.L. No. 101-783, 103 Stat. 183 (1989). Bank United of Tex. FSB v. United States, 50 Fed.Cl. 645, 646-50 (2001) (“Bank United II"). The United States cross-appeals a portion of the trial court’s damages award as well as its interpretation of a regulatory forbearance agreement which formed the basis for its grant of summary judgment as to liability. Bank United of Tex. FSB v. United States, 49 Fed.Cl. 1 (1999), modification denied, 50 Fed. Cl. 327 (1999), (“Bank United I”). Because we conclude that the trial court correctly interpreted the forbearance agreement, we affirm the grant of summary judgment as to liability. However, because we hold that the trial court failed to sufficiently explain the basis for the portion of its damages award challenged by the government, we reverse the trial court’s award of damages as to $3,942,500. Accordingly, we affirm-in-part, and reverse-in-part.

I. CONTRACT INTERPRETATION

As an initial matter, the government contends that the trial court erred in its interpretation of the capital forbearance agreement. The provision at issue provides, in relevant part, that:

The FSLIC will forbear, for a period of ten years following the date of consum*665mation of the acquisition (“Effective Date”), from exercising its authority to take action under Section 563.13 ... for any failure of the Resulting Institution to meet the regulatory capital requirements arising solely from: ... (c) the assumption of the liabilities of the Acquired Institution as of the Effective Date; ... or (2) [Bank United’s] assumption of the Acquired Institution’s regulatory capital deficiency as of the Effective Date; provided that, [Bank United] maintains a ratio of regulatory capital to total liabilities at a level no less than [prescribed by the Capital Plan].

Interpreting this provision, the trial court concluded that, once Bank United assumed the liabilities of the “Acquired Institution,” the forbearance agreement obligated the government to forbear from enforcing its regulatory capital regulations so long as Bank United remained in compliance with the lower capital ratios set forth by the Capital Plan. Bank United I, 49 Fed. Cl. at 5. The trial court explained that, “[s]ince the forbearances ... were designed to enable Bank United to survive the absorption of the liabilities of [the Acquired Institution], it seems apparent that the modified capital standards in the [Capital Plan] are operative at all times, and require no separate showing.” Id. Thus, the trial court held that the forbearance agreement “permitted [Bank United] to meet modified capital requirements in lieu of regulatory capital requirements, for a period of ten years.” 50 Fed.Cl. at 327.

On appeal, the government argues that Bank United’s assumption of the liabilities of the Acquired Institution did not entitle it to operate under the lower capital ratios set forth by the Capital Plan. Rather, according to the government, the agreement provided that Bank United could operate under the Capital Plan’s minimum ratios only during the period between the Effective Date of the transaction and the placement of the $110 million of subordinated debt. Thereafter, argues the government, the capital ratios set forth by the Capital Plan should be read as coming into effect only upon the occurrence of the “triggering events” listed in the agreement. Thus, under the government’s interpretation, Bank United’s “assumption of the Acquired Institution’s regulatory capital deficiency” triggered the government’s obligation to forbear. However, once Bank United came into compliance with the generally-applicable regulatory capital requirements upon issuance of the subordinated debt, “its failure to meet the regulatory capital requirements due to its assumption of [the Acquired Institution’s] regulatory capital deficiency was erased,” thereby deactivating the Capital Plan and requiring Bank United to comply with the generally-applicable regulatory capital requirements unless the Capital Plan was again “triggered.”

Hyperion and Bank United respond that both the language and structure of the agreement support the trial court’s interpretation. According to Appellants, Bank United’s assumption of nearly $5 billion in liabilities as part of the acquisition triggered the forbearance, allowing Bank United to operate under the Capital Plan’s lower ratios during the forbearance period. Hyperion and Bank further point out that the modified capital ratios were incorporated into related “Regulatory Capital Maintenance Agreement” which defines the “Capital Plan Level” as “[t]he level at which [Bank United] is required to maintain its Regulatory Capital pursuant to paragraph one of the forbearance letter.” Finally, Hyperion and Bank United contend that the trial court’s interpretation is further supported by the interpretation of the parties as evidenced by Bank United’s reporting of its compliance with the Capi*666tal Plan rather than the higher generally-applicable capital ratio requirements.

We agree with the trial court to the extent it held that the minimum ratios of the Capital Plan continued to be operative after the issuance of the subordinated debt, and we reject the government’s proffered alternative interpretation under which the Capital Plan is “triggered” and then deactivated by the subordinated debt issuance. Giving the language of the agreement its ordinary meaning and reading the agreement in its entirety, we conclude that the trial court correctly held that the forbearance continued after the issuance of the subordinated debt. There is no suggestion here that transactions occurring after the initial acquisition caused the bank to be out of compliance, although such events might have eliminated the forbearance. The transaction upon which the government relies-the issuance of the subordinated debt-was not an event that caused the bank to be out of compliance, but rather an event that increased the bank’s regulatory capital. The conflict that the government perceives between affording a 10-year term to the regulatory forbearance and the provisions granting forbearance solely on account of the acquisition transaction simply do not exist on the facts of this case. Accordingly, we affirm the trial court’s grant of summary judgment on the issue of liability.

II. LOST PROFITS

“One way the law makes the non-breaching party whole is to give him the benefits he expected to receive had the breach not occurred.” Glendale Fed. Bank, FSB v. United States, 239 F.3d 1374, 1380 (Fed.Cir. 2001) (citing Restatement (Second) of Contracts § 344(a) (1981)). “The benefits that were expected from the contract, ‘expectancy damages,’ are often equated with lost profits, although they can include other damage elements as well.” Id. (citing Restatement (Second) of Contracts § 347 (1981)).

At trial, Hyperion and Bank United argued that they were entitled to expectancy damages of either $558 million or $96 million as measured by two alternative models of lost profits. Both models purported to measure the profits that Bank United would have made by investing the “leverage capacity” it would have enjoyed absent the government’s breaches.

The trial court rejected Appellants’ lost profits claims, finding that the capital infusions from Hyperion allowed Bank United to prevent the government’s breach from negatively affecting its growth or, by extension, its profits. Bank United II, 50 Fed.Cl. at 654 (Appellants “could have and, in fact, did mitigate [the] damages potentially flowing from the enactment of FIRREA.”); id. at 662 (‘We find that not only could plaintiffs have easily mitigated the loss of leverage capacity, they actually did so.”). This conclusion was based in part on two findings by the trial court.

First, the trial court found that “Muring the damages window, the breach impacts did not prevent Bank United from successfully pursuing its forecasted business operations, including retail growth supplemented by wholesale purchases.” Id. at 662. The trial court reached this conclusion by comparing the predicted growth strategy reflected in Bank United’s pre-breach business plans to Bank United’s actual performance. Contrary to the rapid wholesale growth reflected in Appellants’ lost profits models, the trial court found that Bank United’s pre-FIRREA business plans “projected a well-capitalized thrift primarily focusing on retail growth supplemented by wholesale purchases.” Id. at 659; id. at 660 (“The contemporaneous evidence indicates that Bank United pursued a plan, first expressed in the [De*667cember 1988 Holding Company Application] and never contradicted in any subsequent business plan, nor by the actual bank, to first grow its retail base and to supplement its growth with wholesale purchases.”). Thus, the court concluded that Bank United’s pre-FIRREA plans were not consistent with the aggressive wholesale strategy reflected in Appellants’ lost profits models. Further, the trial court found that Bank United’s actual growth was consistent with its pre-FIRREA plans, indicating that the effects of the breach were fully mitigated. Id. at 662.

Second, the trial court found that “Bank United had the financial ability to increase its wholesale assets during the damages window but determined that appropriate purchases were not available.” Id. Thus, the trial court found that the assumption inherent in both lost profits models that Bank United would have invested its leverage capacity in wholesale assets was undercut by Bank United’s actual failure to do so. Id. at 662 (“Thus, even though Bank United had the ability to pursue more investments, including wholesale investments, than it actually did, there were frequently no assets in the market that satisfied its risk and yield criteria.”); see also id. at 663 (“Even though Bank United needed to deploy excess liquidity to maximize income, it was often unable to do so because there was a ‘lack of high quality assets’ available in the market.”) (citation omitted).

On appeal, Hyperion and Bank United contend that the trial court’s conclusion that Bank United would not have used its leverage capacity to engage in extensive wholesale growth absent the government’s breach is clearly erroneous. Appellants point to two supposed flaws in the evidence supporting the trial court’s finding. First, they claim that “the business plans and projections on which the trial court relied are essentially useless in answering the question of how much Bank United would have grown absent the breach.” This is so, they argue, because one of the business plans relied on by the trial court was based on incomplete information about the thrift operation being acquired and the remainder were prepared under the threat of the government’s breach and therefore do not accurately reflect how Bank United was expected to grow in the absence of the breach. Second, Appellants argue that the trial court misinterpreted isolated evidence that Bank United was occasionally under-leveraged, leading it to reach the erroneous conclusion that this was a persistent problem caused by Bank United’s inability to find suitable wholesale assets.

With respect to Appellants’ argument that the business plans relied on by the trial court do not accurately reflect how Bank United would have grown in the absence of the breach, we note that testimony was offered at trial supporting the general proposition that business plans submitted as part of the regulatory process accurately reflect the manner in which managers intend to operate their thrifts. Moreover, although Appellants assert that “[t]he December 1988 business plan was prepared before an agreement was reached on the capital promises and was based on limited information about the thrift operation being acquired,” they do not identify any specific inaccuracies in the business plan that would render the trial court’s reliance on the plan clear error. Similarly, Appellants claim that the remaining business plans relied on by the trial court were prepared under the threat of the government’s breach, but fail to cite any evidence specifically supporting their assertion. The record, however, contains testimony and other evidence indicating that as late as mid-1989, Bank United’s managers expected that the government would honor the regulatory forbearances *668and that various projections and other documents were prepared based on that expectation. In light of this evidence and of the record evidence as a whole, we cannot conclude that it was clearly erroneous for the trial court to find that Bank United’s pre-FIRREA business plans and projections accurately reflected the pre-breach thrift’s expected growth and manner of operation.

As for the contention that the trial court erred in its conclusion that Bank United was often underleveraged during the damages window, Appellants in effect ask this court to re-weigh evidence presented to the trial court and reach a different interpretation. Appellants do not contest the fact that Bank United’s officers and managers referred to Bank United as under-leveraged or having excess liquidity on various occasions, but rather claim that the “trial court ignored the evidence explaining the Bank’s occasional underleveraged position.” We decline to draw our own inferences regarding the extent of Bank United’s underleveraged position from the record evidence. Thus, we conclude that the trial court did not clearly err in finding that Bank United was underleveraged for significant periods of time between 1989 and 1992.

Having rejected Appellants’ specific challenges to the trial court’s factual findings, we similarly reject the contention that the trial court was clearly erroneous in its conclusion that Bank United successfully mitigated the effects of the government’s breach through capital infusions from Hyperion. Bank United II, 50 Fed.Cl. at 662. Accordingly, we hold that the trial court did not err in declining to award damages based on the lost profits models presented at trial. Moreover, because the trial court’s finding regarding mitigation fully supports its decision not to award expectancy damages, we find it unnecessary to address Appellants’ further arguments with respect to lost profits.

III. COST-OF-SUBSTITUTE CAPITAL

Hyperion and Bank United next argue that the trial court erred in rejecting their claim for mitigation damages in the amount of $117 million. This amount is based on Appellants’ “cost-of-substitute-capital” model. This model assumes that Bank United would have needed to raise approximately $228 million dollars in December 1989 to mitigate the loss of leverage capacity caused by the government’s breach. Bank United II, 50 Fed.Cl. at 655.

The trial court rejected this model, explaining that “the most serious flaw in this model is its assumption that all of the lost ratio would have to be restored immediately in December 1989 to forestall economic harm.” Id. “But an immediate infusion was not required to mitigate,” continued the trial court, because mitigation “could have been accomplished, as it was, on a piecemeal basis as and when it appeared to be desirable to take advantage of investment opportunities.” Id.

On appeal, Hyperion and Bank United challenge the trial court’s reasoning, contending that the “piecemeal approach, which supported Bank United’s actual post-breach operations, would have been totally inadequate to mitigate the lost profits from the substantial additional wholesale growth that the Government’s breach prevented.” Thus, argue Appellants, mitigation of the lost profits that would have been earned by investing its leverage capacity would have required that the lost capacity be immediately restored in full after the breach.

Appellants’ arguments regarding their cost-of-replacement-capital model depend

*669on a premise that we have already concluded was correctly rejected by the trial court. The cost-of-replacement-capital model purports to calculate the hypothetical costs of mitigating the profits that Bank United would have earned by investing the leverage capacity it would have had absent the government’s breach. As explained above, however, the trial court properly rejected this theory of lost profits, concluding that capital infusions from Hyperion allowed Bank United to invest in wholesale assets in the same manner and to the same extent that it would have absent the breach. In short, Appellants seek the hypothetical costs of mitigating lost profits that the trial court correctly determined were not, in fact, “lost” at all. Thus, because we have concluded that the trial court did not clearly err in finding that Bank United would not in fact have earned these purportedly-lost profits absent the government’s breach, it necessarily follows that the trial court’s denial of the hypothetical costs of mitigating such profits cannot have been clear error.

IV. MITIGATION DAMAGES

Hyperion and Bank United next argue that the trial court erred in denying their post-trial motions. They contend that the trial court erred by: 1) not awarding them $35,338,988 that was infused into Bank United as equity during 1990 and 1991; 2) refusing to award them $57,074,034 in damages representing a 20 percent rate of return on the $35,338,988 that was infused; and 3) declining to award them $54,126,972 in dividend payments made on the preferred stock issued by Bank United in 1992.

A.

In explaining its calculation of damages, the trial court began by stating that “Plaintiffs are entitled to recover their actual costs incurred in mitigation of the lost leverage capacity caused by FIRREA.” Bank United II, 50 Fed.Cl. at 665. The trial court went on to find that Appellants would have been entitled to the “proven costs” of Hyperion’s infusing more than $35 million into Bank United but that no costs associated with the infusions had been demonstrated. Id.

In post-trial motions, Hyperion and Bank United argued that the trial court’s finding that the capital infusions were mitigating transactions entitled them to recover the $35 million that was infused. The trial court rejected Appellants’ argument, explaining at length that:

It’s important to keep in mind as the starting point of the analysis concerning Plaintiffs’ motions that the portion of the bank’s capital ratio provided by the supervisory goodwill, and by the forbearances which Bank United lost upon the enactment of FIRREA, represented borrowing capacity. The FIRREA breach did not remove one dime of tangible investable money from the asset side of the bank’s balance sheet, ... The new capital, insofar as mitigation is concerned, was replacing borrowing capacity, not tangible assets.

The trial court further emphasized this distinction between “assets” and “borrowing capacity” by stating that the capital infusions “not only restored borrowing capacity, but provided that much cash for the bank’s growth.” In other words, the infusion of these funds not only replaced lost leverage capacity, but also resulted in a net increase in the assets of Bank United in the amount of $35 million. Accordingly, the trial court concluded that “[t]o award the principal amount of such capital infusions would be replacing an investment Plaintiffs still have or have transferred.”

On appeal, Hyperion and Bank United contend that even if the trial court was *670correct in concluding that the capital infusions conferred benefits on Bank United beyond the mitigation of lost leverage capacity, such benefits were both “incidental” and unwanted and therefore should be ignored in calculating mitigation damages. In addressing the distinction drawn by the trial court between the restoration of borrowing capacity and the addition of assets to Bank United, Appellants concede that “[i]t is true that the payment of $35,338,988 into Bank United as common equity capital had two distinct effects.” Appellants explain that the infusion of these funds both “added back some of the leverage capacity lost due to the breach” and “added $35,338,988 in cash to Bank United that otherwise would have had to be borrowed.” Appellants continue by stating that “[t]his [second] effect, the savings in day-to-day funding costs that would otherwise have been incurred, is separate from the restoration of leverage capacity and goes beyond mitigation of the harm caused by the breach.”

Appellants’ characterization of the effect of the $35 million infusions as merely “the savings in day-to-day funding costs” points out the flaw in their argument. The effect of the capital infusions was not merely a savings in borrowing costs. Rather, if the capital infusions replaced funds that would otherwise have been borrowed, they prevented Bank United from incurring a $35 million debt that it eventually would have been required to repay. Thus, the capital infusions not only relieved Bank United of the need to make interest payments on the $35 million that it otherwise would have borrowed, they also relieved Bank United of having to repay principal in the amount of $35 million.

By assuming that the capital infusions merely resulted in a reduction in borrowing costs, Appellants completely fail to take into account the effect on Bank United of having avoided a $35 million liability. Whether the $35 million is assumed to have been invested or used to have replaced funds that otherwise would have been borrowed, the result is the same: Bank United was worth an additional $35 million as a result of the net increase of its assets in that amount. Accordingly, as the trial court found, in addition to mitigating the lost leverage capacity resulting from the government’s breach, the infusion added $35 million in “investable assets” to Bank United. And, as Bank United’s owner, Hyperion necessarily captured that increase either through the increased value of Bank United or as dividend payments.

Appellants’ further argument that this court’s precedents compel a contrary conclusion is similarly flawed. Appellants contend that, as in Bluebonnet Sav. Bank, FSB v. United States, 266 F.3d 1348, 1355-56 (Fed.Cir. 2001), this $35 million capital infusion constitutes a “financing cost” that should be recoverable in whole. Similarly, they contend that, as in Hughes Communications Galaxy, Inc. v. United States, 271 F.3d 1060, 1066-69 (Fed.Cir. 2001), the $35 million should be considered to be part of the additional cost of obtaining a substitute performance or “cover.” At bottom, these cases stand for the well-established principal that a non-breaching party may recover the actual costs of mitigating a breach. As explained above, the trial court found that the $35 million infused into Bank United was never expended or “lost” and therefore was not a “cost” of mitigation.

In sum, Appellants’ argument fads to demonstrate any error, much less clear error, in the trial court’s finding that they are not entitled to recover the $35 million because this amount represents an investment that they “still have or have transferred.” Accordingly, we must decline to hold that the trial court erred in refusing *671to award this amount as mitigation damages.

B.

In their post-trial motions, Hyperion and Bank United claimed that they were entitled to $57,074,034 in mitigation damages representing a 20 percent rate of return on the $35,338,988 that was infused. The trial court rejected Appellants’ claim, finding that “[t]o award a hypothetical rate of return on [the $35 million] would be a duplication of whatever return the infusing Plaintiffs have already received from their investment.”

On appeal, Hyperion and Bank United renew the argument they made to the trial court. Thus, Appellants contend that “[i]f the Government had not breached, Plaintiff-owners would have retained the $35,338,988 for use in other investments and also received the returns actually produced by Bank United without the need to add more regulatory capital.” According to Appellants, any “return” earned as a result of the $35 million that was infused into Bank United represents the mitigating effect of that infusion. Therefore, they contend that the trial court erred in concluding that rewarding them lost returns on the infused funds would be a “duplication” of returns they already received on their investment.

Appellants’ argument again ignores the distinction drawn by the trial court between the restoration of borrowing capacity and the addition of “investable assets” to Bank United. As discussed above, the trial court found that “[t]he new capital, insofar as mitigation is concerned, was replacing borrowing capacity, not tangible assets.” Thus, the capital infusions “not only restored borrowing capacity, but provided that much cash for the bank’s growth.” Accordingly, in addition to the mitigating effect of restoring borrowing capacity, the infusions provided Bank United with an additional $35 million that itself could be invested and on which Bank United earned a return. Bank United and Hyperion fail to demonstrate any difference between the rate of return earned on the $35 million that was infused and the rate of return that otherwise would have been earned on alternative investments. Thus, we have no basis for concluding that Appellants suffered any damages by investing the $35 million in Bank United instead of an alternative investment.

In short, Appellants’ argument fails to demonstrate any error in the trial court’s rationale for refusing to award them supposedly-lost returns on the $35 million that was infused into Bank United. Accordingly, we conclude that the trial court did not clearly err in denying Appellants’ claim for $57,074,034 in lost returns.

C.

Hyperion and Bank United next contend that the trial court erred in declining to award them $54,126,972 in dividend payments made on the preferred stock issued by Bank United in 1992.

In denying Appellants’ post-trial motions, the trial court stated that “[t]he essence of my disagreement with Plaintiffs’ position is that the reimbursements they now seek are not for the transaction costs of mitigating ... but for the money raised to pursue aims which would have required infusions or acquisitions of real cash to accomplish, even in the absence of the FIRREA breach.”

On appeal, Hyperion and Bank United argue that “the leverage capacity restored by the preferred stock’s addition of $85,500,000 in regulatory capital to Bank United would, absent the breach, have been available without the need to pay preferred stock dividends.” Therefore, *672they contend, the trial court erred in refusing to award them the dividend payments as mitigation costs.

As with the arguments addressed above, Appellants fail to demonstrate any clear error in the trial court’s findings. The trial court limited its initial award of damages to the transaction costs incurred by Appellants in the 1992 preferred stock issuance. Bank United II, 50 Fed.Cl. at 665. Again in its ruling on the post-trial motions, the trial court explained that the “essence of [its] disagreement with [Appellants’] position” was that they sought an award that went beyond transaction costs. The trial court’s finding that Appellants’ mitigation costs associated with the 1992 preferred stock offering were restricted to transaction costs is consistent with testimony offered by the government’s lead expert at trial that, because the value of the cash proceeds of a capital offering equal the expected expense of future dividends, the true costs of such a transaction are limited to floatation costs. We have previously declined to hold such a finding clearly erroneous, and we do so again here. See Cal. Fed. Bank, FSB v. United States, 245 F.3d 1342, 1350 (Fed.Cir. 2001).

V. CROSS-APPEAL ON DAMAGES

The issue remaining for us to address is raised by the government’s cross-appeal of the trial court’s award of $3,942,500 representing the transaction costs associated with Bank United’s $85.5 million preferred stock issuance. In its damages decision, the trial court found that Bank United’s issuance of 3,420,000 shares of noncumulative, preferred stock, which occurred on December 29, 1992, was a mitigating transaction. Bank United II, 50 Fed.Cl. at 665. Accordingly, the court found that Appellants were entitled to recover $2,992,500 in commissions and $950,000 is other expenses associated with the stock offering. Id.

On appeal, the government contends that the trial court committed clear error in finding that the 1992 stock issuance was a reasonable mitigating transaction. This is so, accordingly to the government, because “any ‘adverse effect’ of the breach upon Bank United’s leverage capacity ended on December 31, 1992, and, thus, any benefit derived from any ability to operate at a capital level lower than otherwise required disappeared.” Moreover, emphasizes the government, “any adverse effect during 1992 was insignificant and immaterial, because (as of the beginning of that year) the capital level required by the capital forbearance and the prevailing capital requirements converged at three percent, though there were minor computational differences” (citation and emphasis omitted). Therefore, the government argues, the “transaction costs related to raising $85.5 million in equity bore no relationship to the costs necessary to replace [the] insignificant and immaterial amount of lost leverage capacity” remaining on December 29, 1992, two days before the end of the damages window.

Appellants reply to the government’s argument by pointing out that “[t]he trial court did not rule that the preferred stock mitigated ‘two days’ of lost leverage capacity” but rather “focused on mitigation that, in [its] view, enabled [Appellants] to pursue their long-term strategy and plans for Bank United.” Appellants further claim that the timing of the stock offering does not preclude its being a mitigating transaction, arguing that “[b]y its nature, the loss of leverage capacity caused by the breach had continuing, long-term effects on Bank United that did not disappear on December 31, 1992.”

We cannot accept Appellants’ mere assertion that the government’s breach had *673“continuing, long-term effects” which were mitigated by the 1992 stock offering. Appellants have entirely failed to direct our attention to any evidence in the record demonstrating any effects of the breach extending beyond the damages window. Nor is it apparent from the trial court’s extensive decision what such continuing effects might be. Accordingly, we reverse the trial court’s award of $8,942,500 representing the transaction costs associated with Bank United’s 1992 preferred stock issuance.

CONCLUSION

For the foregoing reasons, we affirm the trial court’s grant of summary judgment as to liability and its denial of the additional damages sought by Appellants. However, we reverse the trial court’s award of $3,942,500 in mitigation damages associated with the 1992 preferred stock issuance.

No costs.